Understand the data model of data shares

If you have access to an Atlassian Data Lake connection, you may already be familiar with the data model that’s available in Atlassian Analytics. Learn more about the data model used in Atlassian Analytics.

The data models between data shares and Data Lake connections in Atlassian Analytics will differ. Data shares give you access to all the tables of the Data Lake schema, including pre-transformation tables that aren’t directly available in Atlassian Analytics. These tables provide more flexibility in how you want to model the data for your own use cases.

For Data Lake connections in Atlassian Analytics, we derive many of the tables from the pre-transformation tables to make the data faster and easier to query. Your organization will need to make the extra transformations and calculations on the data models of your data shares to recreate the views used in Atlassian Analytics.

Data freshness

For most tables, it takes at least 30 minutes for changes in your products to reflect in your data shares. However, the jira_issue_enhanced_table table takes about three to six hours for changes to reflect.

Tombstone records

A tombstone record is a row that has a value of true in the envelope_fields_is_tombstone column. This indicates that the corresponding object was deleted from the originating product. For example, if a Jira issue was deleted, you'd see a tombstone record for this issue.

Tombstone records may remain in the data share for up to 37 days from when you delete the object in the originating product.

However, when an object is deleted from the originating product, we immediately purge the content from its corresponding columns in the data share, except for the following metadata columns:

• envelope_fields_generation_counter

• envelope_fields_resource_ari_value

• envelope_fields_version

• envelope_fields_workspace_id_value

• row_refreshed_at

• row_refreshed_at_day

• shard_id

• workspace_id

Read more about these metadata columns.